Search the Community
Showing results for tags 'Motorola 68000'.
Found 4 results
Hi... I just purchased an "Atari ST 520" off of ebay for the first time "Ever"; I never really owned one before so I am a first time of one of these machines. I was wondering has anyone ever programmed Mode 7 style graphics on the Atari ST? I've posted picture of a demo of mode 7 on a Sega Genesis and wonder if anyone ever tried to pull it off on the ST yet? It sounds like a neat little project for somebody.
The interface for a good assembler is just like a text editor, with extra features added to make assembly easier. Take a look at this simulated screenshot, inspired by the Apple ][. This is a multiply routine for the Motorola 68000: There are several things that would make this more of an assembler than a word processor: Under the label "Multiply," there is a blue line stretching across the screen. You could toggle this on or off. Under this line can be shown information about the subroutine (e.g. input/output). Each line of code is indented automatically. The local labels have a period before them, and are not indented. There is a red "+" before the label. Clicking it changes it to a "-" and makes the code disappear. You could click the "-" to make the code reappear. Whether the code is folded or not, it's compiled when requested. When compiled, the branches with the ".s" extension will resolve to a ".b" (8-bit) or ".w" (16-bit) displacement, whichever is the shortest possible. If the extension is left off, assume it to be ".s". That way, you don't have to figure it out yourself. In this example, the screen is 480x360 pixels. Characters are 7 pixels across and 8 pixels down, just like on the Apple ][. In system RAM, this could be handled with one table telling which ASCII character to show (one byte per character), and another table to tell the background/foreground colors for each cell (in each byte, there are 4 bits for background color and 4 bits for background color). By default, the line under labels is enabled, tab width is 8 characters, lines after labels and code automatically indent, and code is not folded. When a mouse is used, the character that the mouse is pointing to is shown in a different color (for example, in the above screen, it would be shown as a white cell with a blue character). Characters would be stored as ASCII. The blue underline is toggled on/off with a control byte, and the tab width is also controlled using a certain byte. You could use any programming language you want, be it 6502, 68K, Z80, BASIC, etc. Regarding the keyboard, there could be additional keys based on what programming language you use. In addition to a regular ASCII keyboard, there could be attachments you could just snap on. For example, a 6502 keyboard attachment might have buttons labeled "LDA," "STA," "CLC," "SEC," "ADC," and "SBC." Next, I'll mention some enhancements you could make to the screen.
I have used a lot of assemblers to program games. I have used Learn to Program BASIC, BasiEgaXorz, and EASy68K. I have also used Apple ][ Basic, C++, and others. There are many different assemblers out there, but what if there was a computer (or maybe an application) with a really sophisticated assembler that could be used for programming games, and other things? The goal is to make programming easier, faster, and more enjoyable. First, I'd like to mention all the essential things that any good assembler needs. Fast interface, as well as fast assembling. The ability to cleanly divide a ROM into sections. Code/data folding. The ability to test code as you write it. Storing colors, but showing them visually, rather than as numbers. Storing graphics for a game as data, and making it show like it would in the program. Compress graphics if necessary. If it's for a system that uses tiles for graphics, computing the mappings for them. Compress data in some way. Test code for length. Being able to make short/long branches automatically according to smallest possible file size. Making sure VBlank code starts and ends properly. For any routine, sort the local labels alphabetically or numerically. Add a number of labels to a ROM that follow a certain character pattern. Add/manage data structures. Lets you pick labels/variables from a list. Calculates a ROM checksum and/or adds code. Pads a ROM to a number of bytes that is a power of 2. I might add more of these. Over time, I will be adding blog posts regarding one or more of these elements. Keep in mind that any images posted in this blog are simulated. The Apple ][ is my inspiration for their look, since it was one of the first computers I grew up with.
This next section is a big one. Wouldn't it be great if you could test code as you programmed it? Well that's where Code-As-You-Go comes into play. The mode can be accessed with a dedicated button on a keyboard. It's labeled "CAYG." Take a look at this: That's the code as you go screen. On the panel at the right, you can enter the data you want to test. On the upper right of the screen is the address that the code will assemble to. In this example, the written code will compile at address $001404. You could instead have it display which line of the source code the code will go in. First, give the subroutine a name. In this example, we have a routine called "TetrisLFSR." This will be a Motorola 68000 version of the NES Tetris RNG routine. The NES version of Tetris iterates its RNG (a 16-bit LFSR) in the following manner: Set the output bit to the XOR of bits 1 and 9, and right-shift that input into the RNG. We will replicate this routine as we enter the code. For this test, enter the input in d0. We need to enter a 16-bit value. Using a mouse, click on the fourth-to-last digit of the d0 register, then type "7259." The digit highlighted in green is the cursor. Note that the register values are displayed in hexadecimal. If you enter an invalid hexadecimal digit, nothing happens. When you enter the last digit, the cursor stays there. (If it were an A-register, the cursor would be red.) Now, time for the first instruction. Type "move.b", tab, then "d0,d2", and hit Enter (if you hit Space, it will tab for you). When you press Enter, the last line of code you wrote is automatically executed in the CAYG window, and its machine language code appears in the window as well. In M68K assembly, the instruction "move.b d0,d2" is represented by $1400. The screen looks like this: Note that after you typed the code line, that instruction automatically executed. The last byte of d0 is $59, so the last byte of d2 is now also $59. The next two instructions are "move.w d0,d1" and "lsr.w #8,d1". These are necessary to retrieve the upper byte of a 16-bit value in d1. After the second line was typed, d1 became $7259. After the third line, it became $0072. In the machine code box is E049, which is the code for "lsr.w #8,d1." Remember, only the compiled code for the last line you typed appears in the machine code box. Next, we want to take the XOR of bits 1 and 9 of the bytes in d1 and d2. Since 1 and 9 differ by exactly 8, no shifting of either byte is needed. Just XOR the bytes by typing "eor.b d2,d1", then pressing Enter. Register d1 is now equal to $2B, which is the XOR of $72 and $59. It is bit 1 from this value we need to extract and get into the X (extend) flag. To do this, type "lsr.b #2,d1", and press Enter. The value in d1 became $0A. But more importantly, look at the X and C flags. They lit up, so their value is 1. Any flag that is clear appears as white-on-black, while a set flag is indicated by the opposite color scheme. Since the XOR of bits 1 and 9 of our 16-bit value was 1, a 1 will be right-shifted in to get the new RNG value. Here is the last piece of the puzzle. Now that we have our output bit in X (and C), we can use a "roxr" instruction to shift it in. Type "roxr.w #1,d0", and hit Enter. And there you have it. The new RNG value is $B92C. With the ability to see the code execute as you type it, coding will become as easy as pie. You could also toggle register updating off/on, and you could also move your cursor to any line in the code, and press a certain button to step through the code and see the results. After finishing the code, press the CAYG button again. All the code you wrote in the CAYG screen will be placed at the place in the source code you were at when you went to this screen. You can then edit it, delete it, or change it as normal. All in all, the code-as-you-go feature could be a breakthrough for future assemblers. No matter whether it's 6502, M68K, Z80, or anything else, it's the next innovation in coding.